MBW Views is a series of op-eds from eminent music industry people… with something to say. The following MBW op/ed comes from Deviate Digital founder Sammy Andrews.
Our robot overlords now have a social network of their own. Moltbook is where millions of AI agents talk to each other about intelligence, efficiency and the future, while humans watch from the outside.
It’s an amusing, if faintly unsettling, prelude to a far more prosaic reality: AI’s real influence on the music business has nothing to do with sentience and everything to do with operations.
Public debate around AI in music remains dominated by creation, ownership and ethics, not least following recent licensing deals. Those questions matter, but they sit some distance away from where advantage is actually being created.
Inside music companies, AI is being applied to forecasting, finance, marketing measurement and commerce infrastructure in ways that closely resemble earlier shifts in banking, retail and travel. The impact is visible not in grand narratives, but in day-to-day decisions that quietly compound.
As an agency, we were early to deliver ‘digital transformation’ audits to clients across the music businesses. Over the past six months, those same organizations have increasingly asked us to run what are now described as ‘AI readiness’ audits.
In the age of agentic search and large language models, structural unreadiness is already costing money and margin across ticketing, merchandise and catalog discovery.
In recorded music, weaknesses tend to surface first in forecasting and planning. Release strategies still lean heavily on recent comparables, editorial signals and territorial intuition. That logic breaks down under catalog scale, fragmented attention and shortened feedback loops.
“The impact is visible not in grand narratives, but in day-to-day decisions that quietly compound.”
Machine-learning models trained on historic streaming behavior, platform mechanics, release timing and audience response are now used to model performance ranges rather than single outcomes. The result is fewer overbuilt campaigns, fewer late corrections, and earlier identification of releases that genuinely compound rather than briefly spike.
Live music has moved in a similar direction. Tour routing, show counts and on-sale strategies increasingly mirror approaches long used in airlines and sports scheduling.
Artist-run businesses tend to adopt these approaches fastest because the commercial feedback loop is immediate. D2C operations now function much closer to modern retail than to traditional music merchandising.
Machine learning is used to forecast inventory, manage fulfillment and model repeat purchase behavior. Production runs are sized with greater precision, reducing underestimates and dead stock. It’s not radical, but it compounds quickly.
If you are a manager and you are not running AI readiness checks across your artists’ businesses, you are already falling behind. The operational surface area of an artist now includes their data, catalog structure, digital properties and marketing systems. Weak foundations here quietly erode leverage over time.
Managers should also be having conversations with their artists about the use of generative AI for assets, video, voice or other applications. Artists hold very different views here, and those differences matter. Avoiding the discussion does not preserve control; it defers decisions to teams, platforms or third parties. Boundaries need to be established deliberately, not by default.
“If you are a manager and you are not running AI readiness checks across your artists’ businesses, you are already falling behind.”
Finance teams and deal-makers have been among the fastest beneficiaries of applied machine learning. Revenue prediction, royalty anomaly detection and fraud identification are well-established in banking and payments. Applied
to streaming, publishing and neighboring-rights data, these techniques surface duplicated identifiers, abnormal payout patterns and unexpected territorial variance far earlier than manual review.
Advance modeling has moved away from static comparables towards multi-scenario stress testing that reflects catalog decay and platform behavior. Judgment remains central, but it now operates within clearer constraints.
In marketing, the big change has been measurement rather than creation. Media-mix modeling, incrementality testing and audience segmentation allow teams to distinguish correlation from causation and reallocate spend accordingly. Generative AI sits within this system as an execution layer in places, producing variants once strategy is set. At the same time, we are already seeing advertising platforms autonomously altering copy and creative without explicit consent, raising real brand-safety and tone-of-voice issues for artists that require firmer guardrails.
Meta has been clear that the volume of AI-generated video inside its ecosystem has increased sharply, which is precisely why it is being separated rather than embedded more deeply into creator workflows. It would be surprising if other major platforms and DSPs did not move in a similar direction as synthetic promotional content continues to scale.
The rapid growth of music business-specific AI products has further muddied the picture. Most amount to little more than snake oil – generic models repackaged for music, often with minimal understanding of rights, accounting or platform mechanics, and built on scraped public data dressed up with glossy UX.
E-commerce and ticketing remain widely under-optimized given their revenue importance. In other sectors, machine learning routinely improves conversion, inventory management and checkout performance. Similar techniques are now being applied to artist stores and ticketing flows, improving on-sale preparation, reducing abandonment and tightening inventory control.
Underlying this is a shift towards AI-mediated discovery and automated decision systems. As consumers increasingly rely on recommendation engines, comparison tools and agents, inclusion depends on structured, reliable data. Artist catalogs, ticketing systems and commerce platforms need to be machine-readable, consistently identified and legally unambiguous. Historic data ownership has never mattered more. Organizations that allowed partners across the business to wall data off from them over time are now paying the price.
AI initiatives often fail not because the models are weak, but because the data they rely on is fragmented, incomplete or internally constrained. AI readiness, in practical terms, means a single source of truth for catalog and product data, stable identifiers across systems, clear territorial and rights metadata, reconciled revenue streams, and governance that allows data to move internally and intentionally.
Outside of the raging wars more widely in audio generation – generative AI warrants only a narrow mention as part of this wider stack. It is now possible for an artist to run an entire promotional cycle without appearing in front of a camera, and some teams are testing this.
Platform responses remain inconsistent. There is no conclusive evidence of systematic down-ranking purely on the basis of synthetic origin, but engagement quality, disclosure and repetition are increasingly policed. How this settles is an operational question rather than a policy one.
Other industries offer a clear precedent. Retailers that delayed investment in forecasting and inventory lost margin and relevance. Financial institutions that lagged on anomaly detection absorbed higher losses. Media publishers that ignored structured data standards were deprioritized by systems they did not control. In each case, the cost was gradual exclusion rather than sudden failure.
The music industry has long prioritized creative output over operational foundations. AI does not alter that imbalance; it exposes it. The organizations seeing durable returns are fixing identifiers, reconciling catalogs, integrating systems, testing aggressively and applying established analytical techniques to familiar problems.
The competitive question is brutally simple. AI advantage in music has nothing to do with automating creativity or predicting hits; it is about who wastes less money, allocates capital more accurately and makes fewer bad decisions, faster.
Treat AI as an overlay and it becomes expensive theater. Fix the foundations and it becomes an advantage your competitors will struggle to unwind.
This article originally appeared in the first issue of MBW’s new premium print publication, Music Business Worldwide Magazine, which is out now.
Music Business Worldwide Magazine is available as part of a MBW+ subscription – details through here.
All MBW+ subscribers get digital access to our new Music Business Worldwide magazine, with six issues released each year. Music Business Worldwide